Linguistic patterns as a framework for an expert knowledge representation in agent movement simulation

摘要

The article shows how the concept of linguistic patterns (LP) combined with expert knowledge can be used to model the dynamics of social groups. Linguistic expressions similar to natural language are the basis for generating virtual forces (VF) that govern the movements of the individual agent. Using logical sentences and providing the methodology to determine the degree of their truth led to the development of rules of agent behavior. Such an approach allows for constructing flexible simulation models. This paper describes this new idea in detail and illustrates its capabilities and properties by simple examples. Furthermore, a series of simulation experiments involving problems of known solutions with stable final configurations is performed and qualitatively analyzed. For additional validation, a suboptimal scattered plot generation method proposed by Drezner is used. A set of simulations for classic problems showed the convergence of agent dynamic behavior in the proposed framework with the solutions provided by the Drezner approach.